The LamarRacing team won the “Best Performance Overall” award at the RoboRacer Competition held as part of the IEEE International Conference on Robotics and Automation (ICRA) 2026. The autonomous race car prevailed against 35 international teams. The foundation of this success is an AI approach that combines data from professional motorsports with human expert knowledge. The result demonstrates how powerful autonomous systems can be created when data-driven learning and human experience work together.
The LamarRacing team at the RoboRacer Competition during ICRA 2026 in Vienna.
AI System Excels on a Challenging Course
The RoboRacer Competition is recognized as an international benchmark for autonomous racing vehicles. This year, the teams had to navigate a particularly challenging course that included a bridge section and deliberately placed gaps in the track boundaries. LamarRacing dominated the time trials and recorded the fastest performance of the entire field. Based on this result, the team received the competition’s “Best Performance Overall” award. In the subsequent one-on-one races, LamarRacing finished fourth.
The students responsible for this success were Lukas Kutsch, Samir Shehadeh, Aftab Akhtar, and Lavinia Kong. The team is supervised by Nils Dengler and Prof. Dr. Maren Bennewitz, with scientific support provided by Sicong Pan. In March 2026, the team published a research paper related to the project.
Formula 1 Data for Autonomous Driving Decisions
A key component of the system is the use of data from experienced Formula 1 drivers. This data was incorporated into the development of an AI model that plans the driving strategy for the autonomous race car. The approach combines large amounts of data with human experience, knowledge, thereby addressing a central challenge in AI research: How can existing experience be integrated into learning processes so that systems can make reliable decisions even in complex and dynamic situations?
Success in the competition provides clear evidence of this under realistic conditions. Instead of operating solely in simulation, the vehicle had to react in real time and make independent decisions on a physical track with changing conditions.
Autonomous Racing as a Testbed for AI Research
The RoboRacer Competition took place as part of ICRA 2026 in Vienna. The conference is one of the world’s most important events for robotics and automation, bringing together researchers and industry representatives from around the globe.
Autonomous race cars serve as a challenging test environment for methods that are relevant far beyond motorsports. Technologies for environmental perception, motion planning, and decision-making under uncertainty are also required in mobile robots, autonomous vehicles, and industrial robotic systems. LamarRacing’s success demonstrates how the combination of data and human expertise can contribute to the development of high-performance autonomous systems.